Bangladesh’s Chars Livelihoods Programme (CLP)

June 2013

CASE STUDY

Pub

lic D

iscl

osur

e A

utho

rized

Pub

lic D

iscl

osur

e A

utho

rized

Pub

lic D

iscl

osur

e A

utho

rized

Pub

lic D

iscl

osur

e A

utho

rized

Pub

lic D

iscl

osur

e A

utho

rized

Pub

lic D

iscl

osur

e A

utho

rized

Pub

lic D

iscl

osur

e A

utho

rized

Pub

lic D

iscl

osur

e A

utho

rized

1

BANGLADESH’S CHARS LIVELIHOODS PROGRAMME (CLP)

Summary

This case study examines Bangladesh’s Chars Livelihoods Programme (CLP) – a large regional social protection and poverty reduction program which aims to secure and promote livelihoods opportunities while at the same time strengthening the resilience of its target population to natural shocks and climate variability. The CLP works with extremely poor households located on fluvial islands or chars in northwest Bangladesh that are particularly vulnerable to annual seasonal flooding as well as random extreme flooding events. The program benefited over 900,000 people during its first phase (2004-2010), and a second phase (2010-2016) is targeting assistance to more than one million people (CLP website, 2012).

The CLP uses a combination of public works, asset transfers (cash/in-kind), livelihoods-related training, market development, and social development activities to achieve its aims. The CLP's key disaster/climate resilience features include: focusing its public works programming on the reduction of flood risks; creating innovative social safety net mechanisms that cushion the program's beneficiaries against disaster impacts; providing post-disaster relief and recovery support to protect and restore the assets/income being built up through the program; and building direct measurement of climate resilience outcomes into its monitoring and evaluation systems. These initiatives have been closely integrated into the CLP's broader livelihoods and social protection focus, creating strong synergies and mutually reinforcing benefits between these areas.

Natural Disaster/Climate Change Context

According to the Intergovernmental Panel for Climate Change (IPCC)’s Fourth Assessment Report, Bangladesh is highly vulnerable to climate change impacts (IPCC, 2007). Bangladesh’s high vulnerability to climate change is due to: its geographic proximity to the Himalayan mountains, source of its rivers and home to climate-threatened glaciers; its flat deltaic topography with very low elevation; its extreme climate variability that is governed by monsoonal rains; its high population density and poverty incidence; and the dependency of the majority of its population on heavily climate-influenced crop agriculture (Ahmed, 2006).

The country is projected to face increased flooding due to glacier melt and more intense monsoons, as well as increased swamping of southern and western coastal areas due to sea level rises (IPCC, 2007). Other potential climate change impact scenarios include variations in the time of the year the monsoon rains appear, with an overall trend towards more frequent rains/flooding and higher climatic variability (Marks, 2010). Many areas are predicted to experience an increased likelihood of drought outside of the monsoon season (IPCC, 2007).

The Jamuna chars which are among the most impoverished and isolated areas of Bangladesh have always been subject to the risk of flooding in the monsoon season (Marks, 2010). The chars are fluvial islands created and destroyed by the erosion and deposition of silt, with many

2

partially or completely submerged during annual floods; they may exist for just a few years or several decades. The char economy is predominantly agricultural, relying on the floods to sustain fertility. At the same time, the flooding leads to the loss of the homes and belongings of many char-dwellers, as well as the loss of access to the cultivation fields on which their livelihoods depend. Char inhabitants living in abject poverty are especially susceptible to the negative impacts of climate shocks (Marks, 2010).

Overview of CLP

The CLP's stated purpose is to improve the livelihoods, income and food security of up to 1 million extremely poor women, children and men living on island chars in the northwest of Bangladesh. This entails the provision of support to extremely poor households that are prone to both natural hazards and economic shocks, such as: flooding; seasonal food insecurity; drought; fluctuations in food prices and the demand for labor. While the CLP does not target climate change per se during the selection and implementation of its activities, it does seek to increase the resilience of its participants to climate impacts (Marks, 2010).

The program works toward this mandate through a combination of activities that are delivered as a package of benefits for extremely poor families (called “core” participating households) over an 18-month period. This includes (Conroy et al, 2010):

• An asset transfer project which provides productive inputs to households to develop and diversify their livelihoods sources and cash stipends to smooth household consumption.

• An infrastructure project which provides employment and protects households from flood impacts (people, livestock, food crops, soil, etc), through public works to raise homesteads. Tube wells with platforms and flood-resistant latrines also are built through the project, and the concept of total open defecation-free communities is introduced.

• Key complementary projects/activities to increase access to markets and services, such as: livelihoods training (e.g., livestock care, fodder production, etc); market development; homestead gardening; primary health care activities and partnership formation to bring basic services like health and education to the chars; and support to micro-enterprises in climate-sensitive areas to access the wider markets of Bangladesh.

• Social capital development via group formation (e.g., formation of village savings and loans groups, social development groups, etc.) and capacity building in areas such as civil rights and laws, self- and mutual-help, disaster management, etc. The wider chars community also participates in these community-wide initiatives, as well as a number of the economic development activities.1

This package of benefits is complemented by a disaster responsiveness strategy which allows the public works/asset transfer projects to respond to the temporary income and asset protection needs created by disasters, e.g., through the provision of employment opportunities, living stipends, emergency grants or in-kind relief (Marks, 2010). The CLP program structure itself has the ability to mount an emergency response through the

1 For detailed information on the comprehensive range of CLP programming, refer to www.clp-bangladesh.org.

3

mobilization of a network of non-governmental implementing organizations and community-based organizers.

All of these projects and activities are designed to be mutually reinforcing in assisting the core households to make a gradual transition out of extreme poverty. The increase in income and assets achieved by the participating households over time gives them a greater capacity for self-recovery from disaster impacts, while the social protection and disaster/climate resilience-building activities provide them with the means to better protect their assets and incomes. This allows the new assets provided to mature and generate new income streams, rather than to be continually eroded by disaster-induced losses (Marks, 2010).

The first phase of the CLP (CLP-1) included the chars of the Jamuna River in five districts of northwestern Bangladesh. It directly targeted 55,000 of the poorest households to receive core support, but also delivered program benefits to other island char residents such as flood protection, latrines, tube wells, and access to savings and credit schemes (Conroy et al, 2010). The CLP-2 has continued to work in three of the districts covered under CLP-1, while expanding into five new districts (see Figure 1), with an estimated outreach to a further 67,000 core households and more than one million island char residents overall (Marks, 2010).

The CLP-1 was funded by the United Kingdom's Department for International Development (DFID) in the amount of GBP 50 million (around US $79 million).2 The CLP-2 is receiving GBP 81.7 million (around US $128.6 million), GBP 70 million from DFID, and around GBP 11.7 million from the Australian Agency for International Development (AusAID) of the Australian government.

2 Using OANDA currency exchange rates as of 29 January 2013.

4

Figure 1. Map of CLP intervention area and size of beneficiary population

Source: Kenward, 2012

5

Program Description

Policy and Institutional Framework

The Government of Bangladesh’s (GoB) 2001 National Rural Development Policy has provided a mandate for the CLP by giving priority to special development programs for disadvantaged regions, including the “char-lands” (Personal communication: S Kenward, CLP, 2012). The 2006 Poverty Reduction Strategy Paper also commits the GoB to halving the country's extreme poverty to 9.5 percent by 2015, a commitment reiterated in the GoB's current five year plan (GoB, 2010); the char dwellers are among Bangladesh's most impoverished people (IMF, 2005).

At the national level, the CLP is sponsored by the GoB's Ministry of Local Government, Rural Development, and Cooperatives (MLGRDC) and executed through its Rural Development and Cooperatives Division. Decisions are taken by a Program Executive Committee (PEC), chaired by the Secretary of the MLGRDC; it includes the CLP project director and a senior management representative, as well as representatives from various MLGRDC divisions and the two bilateral aid donors to the CLP - DFID and AusAID. The PEC is advised by a National Steering Committee and District and Upazila3 (sub-district) level Coordination Committees (see Figure 2).

CLP implementation is contracted out by DFID to a private sector international management agency (IMA), Maxwell Stamp Plc. The program is coordinated by an international and local team recruited by the IMA and its local affiliate. The management staff are based in a Secretariat in Bogra (200km from the capital, Dhaka), and staff teams are located in eight district offices. Each team is responsible for three-five Union Parishads' (local government).

The CLP's regional government (Upazila) counterparts appoint a coordinator to liaise with the IMA. Day-to-day program implementation - such as the hiring of workers and supervision of activities - is overseen by a Project Implementation Committee (PIC). The PIC is comprised of the participating Union Parishads' chair/members and other local inhabitants (Howes, 2006).

Program services are delivered to the chars by local non-governmental organizations (NGOs), called Implementing Organizations (IMOs), under the guidance of the IMA's district offices. Instead of applying a traditional “partnership” model, where plans and programs are either agreed consensually or the local partner submits its own proposal, the CLP has entered into a contractual relationship with the IMOs; the contracts can be rescinded if performance is substandard (CLP/Maxwell Stamp Plc, 2010). Through the IMOs, locally-based Community Development Officers (CDOs) work closely with the participants to deliver many key activities such as the transfer of cash stipends, social development training, and household monitoring. This intensive local support is considered by the CLP to be critical to being able to tailor the program to the real needs of its participants (CLP website, 2012).

3 The districts of Bangladesh are divided into subdistricts called Upazila Parishad (UZP) or thana (Upazilas are similar to the county subdivisions found in some Western countries). Bangladesh, at present, has 500 upazilas. These are the second lowest tier of regional administration in Bangladesh; one level down is the Union Parishads (UPs). The administrative structure consists of Divisions (7), Districts (64), Upazila/Thana, and Union Parishads (CLP website).

6

Figure 2. CLP Organizational Structure and Functions (Information source: Howes, 2006)

Ministry of Local Government, Rural Development and Cooperatives

Program Executive Committee (Decision making role)

National Steering Committee (advisory role)

Chairs PEC District and Upazila Level Coordination Committees

(advisory roles)

CLP Secretariat (Regional Implementing and Monitoring Unit)

Local Implementing Organization (IMO)

Local Implementing Organization (IMO)

Local Implementing Organization (IMO)

Federal Level Local/ Regional Level

7

Key Disaster- and Climate- Resilience Features of the CLP

Infrastructure and Employment Project (IEP)

The IEP represents an investment both in livelihoods and disaster risk reduction. It provides employment and income to its beneficiaries during the seasonal periods of hunger, known as monga,4 while increasing the flood resistance of housing and community structures.

The public works employment scheme focuses on the raising of private individual and clustered homesteads 60 cm above the highest recorded flood water line on earthen plinths (see Figure 3),5 with over 50,000 core and 45,000 non-core households reached during CLP-1 and a further 85,000 expected to be reached in CLP-2 (Marks, 2010; CLP website, 2012). Soil from the adjoining area is lifted and laid by hired labor, then compacted by the recipient families (Howes M., 2006). The plinth’s slopes are planted with fodder grasses to protect them, while providing a supply of feed for livestock (through the asset transfer program). There is sufficient room on the plinths for livestock and to plant a small garden and fruit trees (Marks, 2010). Other flood protection works, which are not a part of the employment scheme, include building or raising/improving existing tube wells for drinking water (4,500 during CLP-1), constructiing sanitary latrines (62,000 during CLP-1) and providing water pumps (Marks, 2010). CLP-2 plans to reach 88,000 households with safe water and improved sanitation (CLP website, 2012).

Figure 3. Raised Plinth

Source: Conroy et al 2010

The recipients of raised plinths are predominantly selected from among the extreme poor, with some slightly better off households included, e.g., where they live in a contiguous block with the extreme poor in the majority; the participation of women in the IEP is ensured by establishing a reserve quota of 35 percent of job cards for women and restricting the working week to five days to allow for rest days and domestic duties (Marks and Vignon, 2009; CLP website, 2012).

4 Monga refers to a seasonal period of low employment/income between September and December each year, between the planting and harvest of paddy, which results in hunger for the poorest/most vulnerable members of the population. Monga can be prolonged and intensified due to factors such as the severity of the near annual preceding floods and monsoon (S Zug, 2006). 5 A plinth is a raised earthen mound with acutely sloping slides

8

The core CLP households are given up to 50 days of work opportunities on plinth-raising activities during monga (Conroy et al, 2010), though some laborers have been recruited from non-beneficiary households.6 Average earnings were around 80-120 taka/day or US $1.00- $1.50 over CLP-1 (see Table 1). Initially, payment was made in the form of a daily attendance allowance; this was later adjusted to payment based on the volume of earth moved, as this has been found to both increase productivity and enable people to earn more. The IMO staff carry out a pre-survey to determine the volume of earth to be moved. The CLP's Innovation, Monitoring and Learning Division (IML) manages a contractor to verify the quantity and quality of a sample of outputs (at least five percent) delivered during the previous month (CLP email communication, 28 January 2013).

Table 1. Overview of CLP-1 Monga IEP Activities, 2005-2009 Year Person Days Participants Average Days

Worked Per Labourer

Average Daily

Wages (Tk.)

Plinths Raised

Plinths Repaired

2005 225,000 14,123 16 80

(Fixed)

0 0

2006 1,016,803 38,000 27 80

(Fixed)

5,729 0

2007 2,634,500 50,750 29 110 10,427 3,500+

2008 755,000 19,371 27 154 6,388 37

2009 821,152 19,706 24 170 6,975 0

Source: Conroy et al, 2010.

A Scheme Implementation Committee (SIC), comprised of around seven members, records the names of those employed, supervises the work, keep attendances and resolves issues that arise during the running of a site (Howes, 2006). SICs are presided over by a Union Parishad representative. Other members can include teachers, imams (religious leaders), social workers and core beneficiaries (Gisby and Kenward, 2010).

The 2007 floods, and subsequent major floods in July 2012, have clearly demonstrated the value of the CLP's disaster mitigation works. Research carried out by the program showed that 500,000 char dwellers took shelter on CLP-raised plinths in 2007 (Marks, 2010). In 2012, 95 percent of both core and non-core households with raised plinths were able to remain in their villages and protect their assets; 18 percent of these also sheltered neighbors and their assets (16 percent). Less than five percent of the core households lost livestock (Kenward et al, 2012).

Most latrines remained intact, with 16 percent of these shared with other households though some were eroded (15 percent) or submerged (17 percent) during the floods. Among the tube wells, 84 percent remained intact. However, only 33 percent of households had access to clean

6 Other flood protection works (tubewells, latrines, water pumps) use a different approach to implementation than the employment scheme.

9

water, due to the lack of a concrete platform for many tube wells, with corrective action recommended by a recent study (Kenward et al, 2012).

The CLP-2 is also undertaking a pilot climate mitigation activity to introduce improved fuel efficient stoves. This is expected to reduce demand and associated environmental costs related to fuel, particularly wood fuel (CLP website, January 2013).

IEP Safety Net (IEP-SN)

Starting from 2007, the CLP-1 introduced a cash grant varying from Tk 800 to Tk 2,000 to those core households unable to participate in the IEP (e.g., those headed by disabled people or women without the physical strength to provide labor. A total of 9,762 households received the IEP-SN during CLP-1 (Conroy et al, 2010) and up to 24,000 households are expected to benefit during CLP-2 (CLP website, 2012). A CLP study found that, during the prolonged 2007 monga season, the recipient households took 40 percent less cash loans and significantly less food loans and credit than control households (Conroy et al, 2010). The participation of disabled-, older- and female-headed households grew progressively over CLP-1 intakes (Hodson, 2009), possibly indicating the success of this and other CLP social equality measures.

Asset Transfer Project

The CLP’s Asset Transfer Project (ATP) is designed to strengthen the income and assets base of its core households, with the longer-term aim of lifting their regular income above the poverty line. The program is comprised of three key elements: the provision of a substantial productive asset; payment of a monthly living stipend; and livelihoods-related training.

The program provides access to a cash grant to the beneficiaries for the purchase of productive assets of their choice. The participants are trained in the selection, care and profitable sale of their assets and in reinvestment strategies to accumulate further assets (Marks, 2010). The level of the asset grant is determined based on the market cost of purchasing and maintaining the chosen item at the time of program implementation (Howe, 2006) and averaged between Tk 13,000-17,000 during CLP-1 (Conroy et al, 2010).

Cattle have generally been the participants' initial asset of choice, whilst a few have chosen to lease land for cultivation or to start up small businesses. This reflects the familiarity of char dwellers with animal grazing as a livelihoods source. The cattle are also easily sold (so can act as a form of savings or insurance for times of need), while providing an ongoing cash flow from the sale of milk, manure or calves. The CLP has further found that Bangladeshi cattle are highly resilient to flood impacts. After income and assets have improved sufficiently, there is substantial evidence that land lease or purchase becomes a preferred reinvestment option (Conroy et al, 2010).

Each beneficiary household also receives a monthly stipend of around Tk 350-600/month (adjusted periodically for inflation)7 to support household consumption and the cost of undertaking the income generating activity. The amount is determined against an estimated baseline income that is obtained for each annual cohort of selected beneficiaries on entry to

7 For the first six months, the household receives two cash stipends: one a livelihood grant of around Tk 250, the other a household support transfer of Tk 350. The household grant was continued for a further 12 months.

10

the CLP. The estimated baseline income of the beneficiary households was around Tk 17.4-20.2 per capita per day in CLP-1 (Conroy et al, 2010).

The stipend is paid to its identified core participant through a passbook for the first 18 months after joining the program (Howe, 2006). The stipend is designed to help households to smooth consumption while the income from the new asset is being built up – i.e., so that households over time feel more secure to save, rather than consume, their income and use it on investment opportunities. IEP employment, especially during periods of shock (e.g., floods, etc.), also is intended to provide a consumption smoothing effect.

Gender Considerations

The CLP’s experience illustrates that decisions to provide assistance in cash or in-kind are context specific and should be informed by consultation with the men and women of the participating communities. For extremely poor female char dwellers, the receipt of the equivalent of up to six-eight months household cash-flow as a lump sum to buy a productive asset was found to carry unacceptably high risks during the first round of asset distributions in CLP-1. Due to cultural limitations on their mobility and a lack of physical security on the chars, local women found the livestock markets to be unsafe. Female household members were also subject to external influence on how to use the funds, e.g., from male relatives.

The program took measures to address these risks including: assisting female participants to make informed choices of assets to purchase through their participation in social development groups and helping them to obtain the assets e.g., IMO staff accompanying the husband to purchase the item, such as cattle (Conroy et al, 2010). The CLP-2's extreme poverty graduation criteria include measurement of women's ability to influence household decisions regarding the sale/purchase of large investments, e.g., cattle (CLP, 2012); this criterion also reflects other program initiatives to increase women's social capital.

Disaster Response

One of the key features of the CLP has been the evolution of its focus on disaster response. In late 2007, the CLP experienced an unanticipated disruption to its implementation when severe flooding struck Jamuna and other parts of Bangladesh, causing widespread damage and loss (Conroy et al, 2010). Following requests from the GoB and DFID, the CLP-1 mobilized a relief response for the char residents in all five program districts. The response was delivered through its IMO/CDO network, including the rescue of people/livestock and the provision of food aid/relief supplies to 128,000 households across 713 villages (Marks and Islam, 2007).

The losses of cash and food crops, and temporarily diminished employment opportunities, also raised concerns regarding potential income and asset erosion. To avert the use of negative coping strategies (e.g., distress sales of livestock, taking out private loans at exorbitant rates, etc), and in response to a request from the GoB, the CLP-1 expanded its IEP activities by one million person days and into three further districts (Marks and Islam, 2007).

Additionally, due to the delay caused by the floods to the annual programmed start of the regular IEP, there were fears about the potential disruption to cash flows and household consumption for the participating core households. The CLP-1 introduced an “IEP advance”,

11

offering every registered worker a Tk. 500 (US $6) advance against future wages. Over 97 percent of CLP-1 households used the advance to purchase food, with on average 80 percent of the advance being spent on food. This strategy was successful in smoothing consumption and off-setting the need to draw down on assets during a time of shock (Conroy et al, 2010).

Further to this, the CLP-1 distributed more than 15,000 erosion grants to core households of between Tk 2,000 - 3,000 (US $24-36) and 24,000 households received roofing materials. Based on the success of this, the CLP-2 has made available over 18,500 grants of Tk 2,000 each in cash or in kind for a variety of emergencies such as homestead erosion, cold periods, cyclones, droughts and fires (Conroy et al, 2010; CLP website, 2012).

Overall, the CLP's experience has illustrated the importance of incorporating climate-resilience oriented social protection activities into programs in locations with high vulnerability to natural hazards and climate change impacts. Studies and evaluations have demonstrated that both the disaster prevention and response actions taken by the CLP have protected the investments being made in livelihoods strengthening and likely saved some lives, including the effective use of a pre-existing decentralized delivery structure. This experience is also reflected in the high beneficiary satisfaction levels recorded after the 2007 flood response (Marks and Islam, 2007).

Key Program Features of the CLP

Targeting and Verification Process

The CLP-1 was initially faced with difficult choices regarding the selection of households for inclusion in the program. Bangladesh's char dwellers have the lowest per capita income in the country (Tango International, 2006), with the vast majority living well below the poverty line. The funds available for the CLP-1 were not sufficient to provide a full package of support to all char inhabitants. For this reason, the CLP-1 targeted first the poorest sectors of char society that were thought to represent the bottom ten percent of the overall Bangladeshi population (Conroy et al, 2010). Consequently, the geographic scope of CLP-1 became focused on the island chars of the Jamuna area. Within these island chars, a core group of the extreme poor were identified to receive a full package of assistance, while simultaneously undertaking supporting interventions to improve the living conditions of all inhabitants.

The CLP follows four steps to identify the core beneficiaries: 1) Geographical targeting of the char dwellers in ten fluvial districts of northern Bangladesh8, identified as among the poorest, most isolated, and most vulnerable to flooding and climate change impacts; 2) Participatory wellbeing analysis, where field staff from the IMO meet with village members and use participatory rural appraisal tools to identify the poorest members within households; 3) Application of CLP Selection Criteria and Validation, where Community Development Officers visit the poorest households, identified by the community itself, to collect household information and match it to the CLP selection criteria; and 4) Verification of eligibility through visits by senior CLP staff to a random sample of at least five percent of the selected households (CLP/Maxwell Stamp Plc, 2012a).

8 CLP-1 covered five districts, while CLP-2 currently covers eight districts.

12

The CLP-1 opted to use the DFID definition of extreme poverty: households with income of less than Tk 19 per person per day (in rural Rajshahi Division). In order to determine which households were likely to meet this criterion, a set of proxy indicators for household income and well-being were developed. These were designed to be as simple and unambiguous (not judgment-dependent) as possible to minimize inclusion/exclusion errors and potential disputes over selection. They were fine-tuned after the first phase of asset transfer during CLP-1, with the criteria described in Table 3 in use since 2006 for CLP-1 and CLP-2 (CLP/Maxwell Stamp Plc, 2012b).

Table 2: CLP’s targeting selection criteria

Criteria Definition 1. Char Household Resident for at least 6 months in a village which has been

classified by CLP as an island char 2. Landless Absolutely zero decimals of land ownership including homestead

land, and having no access to agricultural land including share cropped land and land to be inherited under Bangladesh law. Households renting homestead land are still eligible

3. Assetless Selected households may not own more than Tk 5000 of productive assets. These are mostly comprised of livestock.

4. Credit-less Have no loan outstanding from any microfinance or credit program 5. Jobless Having no salaried occupation and not receiving cash or asset

grants from any other asset transfer program 6. Participation Are willing to attend weekly group meetings, participate in a

livelihoods program and show how the asset shall be cared for. Source: CLP/Maxwell Stamp PLC, 2012b and CLP staff email communications

The beneficiary selection process is carried out by IMO staff who collect data from each proposed cohort of households prior to their entry into the program. The IMOs visit all households, hold group consultations and submit lists of households who they assess as meeting the selection criteria. These households are included as core participants – on average 30 percent, and sometimes more than 50 percent, of village populations. Within the selected households, the vast majority of those chosen to directly receive asset transfers (the core household recipient) are female members.9 Women are viewed by the CLP as central participants in the development process; their availability to attend compulsory weekly social discussion meetings, often greater than that of male household members, has also been a key factor in their selection. At the same time, other male/female members of the household are also invited to participate in various program activities (Conroy et al, 2010).

The CLP's Management Secretariat carries out a verification process in which staff teams revisit and re-interview at least five percent of the selected households to confirm their eligibility. This includes senior staff who are obliged to spend up to one week in the field undertaking verification work. The process is used to exclude ineligible households and to identify eligible households that have been missed. If the targeting errors on either measure exceed five percent, the IMOs have to repeat the entire and onerous selection process (Conroy et al, 2010) followed by re-verification by the CLP.

9 Even though most participating households have a male head of household

13

Targeting Errors

Extensive exclusion errors were identified following the first round of beneficiary selection during CLP-1 in 2006. While the IMOs had followed the selection criteria, the households identified for inclusion tended to be younger and healthier with most having working male household heads. It became apparent that the IMOs had prioritized households that they felt had the most economic potential. Such selection bias may have been justified by the IMOs in terms of trying to maximize the program's benefits – by providing assets to those households able to make the best use of them. Alternatively, it may have reflected an institutional financial incentive to prioritize those households that the IMOs believed would most likely become future customers in their microcredit programs or which already had loans (Hodson, 2009).

The verification process identified these errors and rectified them in 2007. Verification required approximately 450 days of inputs by Secretariat staff over four selection rounds during CLP-1. However, the process demonstrated to the IMOs that the CLP management was serious; as re-doing selection processes entails substantial costs for IMOs, this discouraged attempts to overestimate participant numbers and significantly reduced targeting errors (Conroy et al, 2010; CLP/Maxwell Stamp PLC 2012a). No IMO has had to repeat the selection process more than once, and the selection standards remained consistently high after the lengthy revisions which had to take place during the early stages of CLP-1 (Hodson, 2009). The inclusion error rate of the first CLP-2 intake was only 0.4 percent, and a 2011 CLP poverty assessment found less than one percent of eligible households may be missed due to exclusion error (CLP/Maxwell Stamp PLC 2012a).

Beneficiary Registration and Enrollment

The registration of the core CPL participants is carried out by the IMOs, once the selection and verification process is complete. Beneficiary lists are compiled for each program location, with every household assigned a unique reference number during registration. In 2008, the CLP decided to also assign each household plinth a unique GPS location that can serve as a second reference number for the household (Hodson, 2009).

Following registration, baseline household income and expenditure data are collected by trained IMO staff using a simple questionnaire. This information is then integrated into the core beneficiary database of the CLP Secretariat in Bogra. Changes in household income and expenditure are monitored and recorded over time, along with changes in asset values, food security and nutritional status and other key socioeconomic poverty indicators, such as water and sanitation status and female empowerment. This database has been used to monitor progress and as a sampling frame for measuring the CLP's results and impact. Initially, all core participating households were surveyed monthly after registration, but because of the costs and the difficulty of supervising such large-scale data collection, only representative samples of households have been surveyed since 2008 (Hodson, 2009).

Payment/Distribution Systems

Originally, the CLP used points of sale at particular locations on specific days to disburse payments through the IMOs. While this system has operated satisfactorily, the CLP has initiated a pilot of a new system of direct cash transfers to recipients for their monthly stipend payments

14

through mobile phones in order to reduce both the transaction costs (for the providers and recipients) and potential misappropriation by intermediaries. The CLP has engaged bKash, a Bangladeshi mobile banking service provider, to provide this service.

The participating households have registered for the service, received individual mobile phone SIM cards and opened electronic accounts with bKash. Account holders are able to receive, transfer and cash money through bKash agents. The CLP’s IMOs will transfer the stipend payments directly to the bKash accounts of registered participants, who can then collect their cash from a local registered agent. The pilot covered more than 10,000 households as at January 2013 (CLP email communication, 28 January 2013). If successful, the system will be extended to all 30,000 CLP stipend recipients.

The CLP is among the first development programs in Bangladesh to use mobile cash transfers. Some of the identified benefits for the recipients include: no longer needing to attend a CLP-staffed pay point on a certain day to receive their money, confidence that their accounts will be credited on a specific date, and the ability to access their cash when convenient. Mobile cash transfers also have other potential advantages such as: use of the facility to pay for phone, SMS and other social/trade cash transfer functions (e.g., for temporary char out migrants to send money to relatives); disbursing government social safety net payments; and increasing SIM card and mobile phone ownership through the presence of Kash agents (CLP website, 2012).

For the IMOs, mobile transfers can reduce the risks associated with staff transport of cash to program sites. The system may eventually expand to the payment of IEP wages and emergency grants. Mobile transfers further offer good potential to provide funds quickly to households during times of crisis, often a key factor in the speed and level of disaster recovery.

Monitoring, Accountability and Learning Systems

Monitoring and Evaluation

Not only does the CLP place a strong emphasis on “piloting, learning and adapting” (Conroy et al, 2010) – having an Innovation, Monitoring and Learning (IML) Division in place and strategies to develop and communicate best practice – it also pays specific attention to climate resilience in its monitoring and evaluation (M&E) system and extreme poverty graduation criteria.

The CLP has been implemented incrementally, commencing with small-scale pilot interventions. As the program has learned what works most effectively, and what does not, it has adjusted its approaches and then scaled up coverage. This has helped the program to minimize costly errors and refine program delivery methods to be more responsive and appropriate to beneficiary needs and preferences. Some examples include the tightening up of household targeting, asset transfer and fraud prevention procedures during CLP-1 (Conroy et al, 2010).

The IML Division also plays an internal quality assurance role, regularly documenting the program's progress/impacts, identifying lessons and feeding these back into management; monthly IML-led monitoring was in place during CLP-1 (CLP website, 2012). The CPL's logframe and comprehensive beneficiary database form the foundation of this work (Conroy et al, 2010).

The IML Division uses independent contractors to verify its outputs, outcomes and impacts. For instance, up to 10 percent of the outputs reported by IMOs were independently monitored

15

during CLP-1, with the contractors reporting directly to the IML Director. This approach has been demonstrated to improve IMO performance, ensure households receive their entitlements, and allow issues to be dealt with as they occur. Anonymous and random surveys are also undertaken annually by the contractors to gauge participant satisfaction and the household level impacts of the interventions (CLP/Maxwell Stamp PLC, 2010).

The CLP specifically and systematically monitors the outcomes, impact and sustainability of its initiatives to build climate-resilience among the char dwellers. This is a notable feature of its M&E system, not commonly found in social protection/social safety net programs. Climate- and disaster-sensitive performance criteria have been developed to carry out these assessments, including the incorporation of a flood resilience criterion into the CLP-2's graduation criteria (CLP, 2012). Some examples of this work include reviews/studies of the CLP's emergency transfers/public works following major floods in 2007 floods and of the public works after major floods in 2012 and a recent study on the lifespan and occupancy status of CLP-raised plinths.

Grievance Mechanisms

As outlined under targeting, the CLP follows a rigorous and intensive beneficiary selection and verification process. However, if a household has not been included but believes it should have been, they may contact CLP staff present in the village, who will then follow up on the case.

Accountability and Fraud Prevention

According to CLP reporting, the international management agency and most IMOs have worked satisfactorily in the field. However, a small number of the Union Parishads (UP) engaged in processing disbursements were found to be involved in financial mismanagement or corruption. A 2006 CLP-1 survey reported that around nine percent of respondents had to pay “kick-backs” of Tk. 10 or less from their daily wages on UP managed schemes. In 2007, 25 percent of participants questioned during a customer satisfaction survey felt they had not been paid the correct wages, the majority working on UP-managed schemes; the CLP commissioned an affidavit survey and submitted the evidence from the survey to local government and IMOs. In 2011, the chair of a UP was found to have diverted IEP plinth-raising funds to his own personal business. Complaints by program participants triggered an investigation by the CLP-2, with the funds eventually returned (CLP/Maxwell Stamp PLC 2012b).

The CLP takes a “zero tolerance” approach to corruption, adopting a policy of cancelling IMO contracts and withdrawing future funding from UPs that do not take action against proven cases of corruption. As a result, IEP funding through local government – as compared with IMOs, fell from 74 percent in 2006 to only 26 percent by 2008. However, bypassing the local governments in favor of NGOs is only a short term solution and can lead to backlashes by local governments. The CLP is addressing this issue by training key staff in the Upazilas to manage funds more effectively (CLP/Maxwell Stamp PLC, 2012b) and by training Union Parishad staff (locally elected council members) in their civic responsibilities.

16

In addition, although data from previous CLP surveys suggest that the schemes implemented by

IMOs experience less corruption than those operated by the UPs, some incidences of corruption have been reported in IMO-managed schemes. In 2008, an alternative implementation method was piloted through five IMOs to combat this problem: direct household management of IEP plinth construction. The CDOs directly identified suitable households to be involved in scheme management, using two criteria: (i) ability to identify the poorest and most needy households nearby to participate in the IEP, and (ii) ability to manage payment to laborers without demanding kickbacks (CLP/Maxwell Stamp PLC, 2012b).

The CLP's beneficiary communication channels provide another avenue to reduce leakage, by informing participants of their entitlements. For example, a paid day of induction training is provided for all IEP plinth-raising participants which sets out the conditions of employment. For the ATP, orientation sessions explain the asset selection/purchase process and beneficiary entitlements. In 2008, the CLP-1 also reinforced the independent verification of inputs, regardless of whether managed by UPs or IMOs, by erecting signposts at worksites with a mobile phone number for further information and complaints (CLP/Maxwell Stamp Plc 2012b).

Another effective fraud/leakage prevention tool used by the CLP is the customer satisfaction survey (CSS). The CSSs are undertaken by an independent contractor, who anonymously interviews the laborers employed on infrastructure work, and the beneficiaries of both infrastructure and asset transfer work. This provides a safe environment for the participants to report if they have experienced any threats and demands for kickback payments by those involved in managing the work. The CSS is a rolling survey, with travelling teams of enumerators performing unannounced visits to work sites and conducting interviews. The teams aim to report the information gathered quickly to the CLP, which is able to investigate issues within one to three days. Surveys with workers are also conducted after work is completed, as demands for payments may also be made at this time.

The result of all of these measures is impressive. For example, the levels of reported leakage dropped from highs of 19 percent early in CLP-1 to less than one percent in its final two years (CLP/Maxwell Stamp PLC 2012b).

Communications

The CLP emphasizes giving both its core beneficiaries and the wider island char communities a “voice” in the program, with a view to laying the foundations for char dwellers to develop greater influence over government policies and service delivery. A key activity has been the formation of self-help groups with core beneficiaries and holding of weekly social development meetings; these groups are usually comprised of up to 25 female household members, though some male groups also are formed. At a broader level, Village Development Committees are a key liaison point and have been created where they do not exist (Howe, 2006).

Both structures are used as channels to share information on the program's progress. They have played an important role in the continuous improvement of the CLP, e.g., the refinements made to the ATP delivery methods during phase 1 (Conroy et al, 2010). Regular two-way engagement of the char dwellers – through the community facilitators, groups/committees,

17

customer satisfaction surveys, etc. – has further enhanced the transparency and accuracy of beneficiary selection, along with the accountability of the CLP's service providers.

In terms of communication with external audiences, all findings of CLP research are published by the IML Division on the CLP's website and key learning is disseminated through workshops, primarily with CLP staff. The CLP also participates in a number of working groups which share the findings from other poverty reduction programs in Bangladesh. The IML Division recently established a Communications Unit (2011) to strengthen internal and external communications.

Program Outcomes and Sustainability

The CLP has achieved significant poverty reduction outcomes.10 Savings grew from one-two percent to 20-30 percent of total household expenditure over the life of the program. Substantial increases were found in the value of average household level productive assets, with no further assistance from the program, as well as increases of up to 250 percent in non-productive assets: eg the 2006 ATP beneficiary intake increased their asset base from an average of Tk. 2,024 at entry to Tk. 46,024 by the end of 2009. Income levels were higher, on average, than the national extreme poverty line. By September 2009, the average household income from the 2006 and 2007 intakes had increased by 66 percent and 59 percent respectively. There was also evidence of income diversification among at least 10,000 – 15,000 households in the form of milk and poultry sales, another important factor in reducing vulnerability to climate-related impacts (Conroy et al, 2010).

In addition to the demonstrated contribution of the CLP's flood mitigation works to saving lives and property, a recent study has shown their resilience in actual major flood conditions. Most of the raised latrines and tube wells remained intact during the 2012 floods. The majority of plinths also remained intact (65 percent), with 29 percent partly eroded and only eight percent submerged (Kenward et al, 2012). The overall rate of erosion of the plinths is in line with their calculated lifespan of 15 years (equivalent to the estimated lifespan of an island char). Among CLP-1 core participating households, 74 percent are still residing on their plinths; the remaining 26 percent are not residing on their plinths mainly due to erosion (10.4 percent), unforced relocation (7.8 percent) and eviction (3.2 percent) (Kenward and Islam, 2011). There is room for further improvement, and the CLP-2 is following up on the findings of this research.

A sustainability scorecard, using poverty graduation criteria, has been developed to measure the progress of the CLP2 beneficiaries. The progress of the CLP-1 households will also continue to be monitored in order to provide a longitudinal assessment of the sustainability of program outcomes (CLP website, 2012). From a climate resilience perspective, it will be important to assess whether the resilience of core households has increased to the level where they can absorb the impact of future hazard shocks without a lasting major erosion of their livelihoods.

Lessons Learned

• Regularly recurring natural hazard impacts (e.g., floods) pose a serious threat to the achievement of social protection and poverty reduction objectives. The integration of

10 All figures are taken from CPL1 progress and impact studies by Conroy and Marks (2008), Marks and Sultana (2009) and Scott and Islam (2010) as recorded in Conroy et al, 2010.

18

specific activities to build climate resilience into the CLP's design has brought direct demonstrated benefits to its participants and contributed to the success of its core livelihoods strengthening focus – by protecting the asset and income streams being generated from erosion.

• Targeting specific social safety net payments to protect the income and assets of very poor and vulnerable households from climate-related shocks yields positive results.

• The flexibility and adaptability built into the design of the CLP have been key factors in the development of innovative social safety nets, such as the IEC pay advance, that can respond to both predictable and unanticipated natural disaster and economic shocks.

• The existence of a pre-established and decentralized social safety net system, with a strong community presence, can greatly facilitate a quick and effective relief and recovery response to natural disasters and other shocks.

• The emphasis placed on robust and systematic monitoring and evaluation, community participation and iterative learning approaches – including adequate budgeting for these design features – has made a significant contribution to the high quality of the program. The use of cash or in-kind approaches, or a combination of both, to assist households to build up productive assets needs to be based on robust assessment of the socio-cultural and security context, including gender differences in access to and control over resources.

References

Documents downloaded from CLP website (www.clp-bangladesh.org) except where noted with an *

Ahmed, A. U. 2006. Bangladesh Climate Change Impacts and Vulnerability. A Synthesis. Comprehensive Disaster Management Program. Government of the People’s Republic of Bangladesh.

Chars Livelihood Programme (CLP). Undated. 'The CLP's Graduation Criteria.'

CLP/Maxwell Stamp Plc. 2012a. The CLP’s Approach to Selecting Core Participant Households. CLP/Maxwell Stamp/Australia Government Aid Program/UKAID.

CLP/Maxwell Stamp Plc. 2012b. The CLP’s Approach to Reducing Leakage. CLP/Maxwell Stamp/Australia Government Aid Program/UKAID.

CLP/Maxwell Stamp Plc. 2011. Chars Livelihoods Programme – Bangladesh. Final Report. CLP/Maxwell Stamp/Australia Government Aid Program/UKAID.

CLP/Maxwell Stamp Plc. 2010. Review of the CLP’s Selection and Graduation Criteria. CLP/Maxwell Stamp/Australia Government Aid Program/UKAID.

Conroy, K., Goodman, A.R., and Kenward, S. 2010. Lessons from the Chars Livelihoods Programme, Bangladesh (2004-2010). Paper presented at the CPRC International Conference, University of Manchester. CLP/Maxwell Stamp Plc/UKAID.

Conroy, K., Islam, R., and Marks, M. 2008. The Impact of the CLP 2007 Infrastructure and Employment Programme. CLP/Maxwell Stamp/DFID/Government of Bangladesh.

*Cruz, R.V., H. Harasawa, M. Lal, S. Wu, Y. Anokhin, B. Punsalmaa, Y. Honda, M. Jafari, C. Li and N. Huu Ninh, 2007: Asia. Climate Change 2007: Impacts, Adaptation and Vulnerability. Contribution of Working Group II to the Fourth Assessment Report of the Intergovernmental Panel on Climate Change. M.L. Parry,

19

O.F. Canziani, J.P. Palutikof, P.J. van der Linden and C.E. Hanson, Eds., Cambridge University Press, Cambridge, UK, 469-506.

Gisby, L. and Kenward, S. 2010. 2009 Infrastructure and Employment Programme Worker’s Satisfaction Survey. CLP Innovation, Monitoring, and Learning Division. CLP/Maxwell Stamp/Australia Government Aid Program/UKAID.

Goodman, A.R. and Scott, M. 2010. Achieving Impact – Crucial Factors of Design and Implementation. Paper presented to: Regional Conference on Sustainable Livelihoods and Rural Development in New Delhi, India. CLP/Maxwell Stamp Plc.

GoB (Planning Commission, Ministry of Planning). 2010. Sixth Five Year Plan FY2011-FY2015: Accelerating Growth and Reducing Poverty. Dhaka: GoB.

Kenward, S., Cordier, L. and Islam, R. August 2012. Chars Livelihoods Programme: A study to assess the performance of CLP raised plinths, low cost latrines and access to clean water during the July 2012 flood. Dhaka: Maxwell Stamp Plc.

Kenward, S., Islam, R. June 2011. 'A study to assess the lifespan and occupancy status of CLP raised plinths.' Dhaka: CLP (www.clp-bangladesh.org).

Hodson, R. 2009. Chars Livelihoods Programme. Reflections on the on the CLP Approach to Reducing Extreme Poverty. The Story Continues. CLP/DFID/Maxwell Stamp Plc.

Howes, M. 2006. An Introduction to the Chars Livelihoods Programme.

International Monetary Fund. November 2005. Bangladesh: Poverty Reduction Strategy Paper. Washington DC: IMF.

Marks. Dr. M. 2010. The Chars Livelihoods Programme: Alleviating Poverty and Building Climate Resilience of the Poorest Families. CLP/Maxwell Stamp Plc/UKAID.

Marks, M. & Sultana, T. 2009. Economic Impact of Cattle Transfers during the CLP’s Asset Transfer Programme: 2006 –2008. CLP/Maxwell Stamp Plc/UKAID.

Marks, M., and Islam, R. 2007. The CLP Flood Relief Activities (August 2007): Summary of Relief Activities and Customer Satisfaction Survey, ibid.

Marks, Dr. M. and Vignon, C. 2009. Chapter 12: Char Life, Livelihoods and Development in Floodplain Bangladesh. In: Disasters, Development and Resettlement in Bangladesh: Issues, People and Policies. Mohammad Zaman, K. Maudood Elahi, Hafiza Khatun eds. Draft.

TANGO International/World Food Programme (WFP). 2006. Rural Bangladesh Socio-Economic Profiles of WFP Operational Areas and Beneficiaries. Dhaka: WFP.

Email exchanges and phone interviews with:

Stuart Kenward, Director, CLP Innovation, Monitoring and Learning Division.